Using the Analysis of Logistic Regression Model in Auditing and Detection of Frauds

Abstract

Fraud is defined as intentional actions in which one or more people, including from the management, employees, or the third parties, venture to obtain an unjust or illegal benefit. According to the researches, the average cost of fraud was determined as 5% of total incomes. The fraud, which has the results like a financial iceberg besides the direct losses, causes damages like loss of reputation, and adverse effects of customer relations. Auditing and detection of fraud, which has such vast effects, is of great importance.
In this study, we have developed a model that is designed for detecting mistreatments with logistic regression and the abuses in the performance-based salary system in the health sector. For this, some imaginary surgery data were added into the actual data of laparoscopic cholecystectomy operations performed in a public hospital in 2015, and to distinguish this fictitious data, the success of the generated logistic regression model was tested. Consequently, it shows that the model had 83.30% of the success rate for detecting the false data added to real data.